Dense Statistics on Cortical Thickness and Myelin Reveals Adolescent Brain Development

Stand-By Time

Tuesday, June 27, 2017: 12:45 PM  - 2:45 PM 

Submission No:


Submission Type:

Abstract Submission 

On Display:

Monday, June 26 & Tuesday, June 27 


Dongjin Kwon1,2, Adolf Pfefferbaum1, Edith Sullivan2, Kilian Pohl1


1SRI International, Menlo Park, CA, 2Stanford University, Stanford, CA

First Author:

Dongjin Kwon    -  Lecture Information | Contact Me
SRI International|Stanford University
Menlo Park, CA|Stanford, CA


Studies on cortical brain development, such as the National Consortium on Alcohol and NeuroDevelopment in Adolescence (NCANDA), commonly confine image analysis to regional scores derived from pre-defined brain parcellations, called atlases. This approach simplifies analysis by reducing high dimensional image data to a relatively small number of measurements yet risks missing meaningful developmental patterns whose boundaries do not align well with pre-determined parcellations. Herein, we provide an example to discover brain developments by computing spatially dense statistics on cortical thickness and myelin across the NCANDA sample.


The example was based on the NCANDA baseline structural MRIs acquired on Siemens 3T TimTrio in 226 adolescents (108 male and 118 female; age 12-21 years). Each MRI was denoised, inhomogeneity-corrected, and skull-stripped [7]. FreeSurfer [1] was applied to each skull stripped T1w MRI resulting in a surface mesh of the brain, which was then refined by aligning the T2w to the T1w data. The surface mesh was mapped onto the Human Connectome Project (HCP) template (i.e., 2mm standard CIFTI grayordinate space [5]) by aligning the cortical folding [8]. For each of the 64k vertices of the template mesh, we measured each subject's cortical thickness (corrected for surface curvature) and cortical myelin content (based on T1w/T2w ratios [4,5]). For each vertex and score type (cortical thickness or myelin), sex and ethnicity was regressed out of the associated scores via a general additive model, and a linear model was robustly fitted to the residual scores to obtain age slopes. For each score type, we assessed the average scores and the age slopes on the 3D inflated template and its 2D flattened map overlaid with atlases of Desikan-Killiany [2] and Destrieux [3] from FreeSurfer, and HCP multi-modal parcellation (HCP_MMP 1.0) [6].


Fig. 1 shows the average cortical thickness (a), myelin (b), and corresponding age-related differences (c, d). As reported in [4], cortical thickness was inversely related to myelin in overall distribution except for primary motor cortex and frontal pole. In general, teenagers showed cortical thinning with age, except for the precentral gyrus, which was thicker with age. Evidence for greater myelin occurred with age in the precentral gyrus but less so in precentral sulcus. Along precentral gyrus, both scores showed age-related increases especially in the dorsal area. Occipital cortex regions showed complex aging patterns for both measures.
Fig. 2 displays the 'flattened' version of the left hemisphere with three different parcellations overlaid. Of the three, the Desikan-Killiany (a) was the coarsest resulting in regions containing de- and increasing aging patterns (arrows around precentral cortex). The Destrieux (b) defined finer parcels, which was better in separating patterns in opposing directions; however, the boundaries of the atlas did not closely fit with the patterns (arrows around precentral gyrus). The HCP_MMP (c) best aligned with the aging patterns in the precentral cortex; however, this region also included the paracentral lobule (arrow), which did not reveal strong aging patterns. Age-related size declines resulted in poor boundary matches, notable in the occipital cortex, for all three atlases.
Supporting Image: OHBM2017_abstract_djk_Fig1.png
   ·Fig. 1. Inflated surface views for average cortical thickness (a), myelin (b) and their changes per year (c, d). Top 2 rows show the left hemisphere and bottom 2 rows show the right hemisphere.
Supporting Image: OHBM2017_abstract_djk_Fig2.png
   ·Fig. 2. Flat surface views of the left hemisphere for changes of cortical thickness (top) and myelin (bottom), overlaid with Desikan-Killiany (a), Destrieux (b), and HCP_MMP 1.0 (c) atlases.


As we show with respect to cortical thickness and myelin scores of this NCANDA sample across the adolescent age range, recent advances in the brain surface registration enables accurate and dense statistical analysis of brain morphometry. This dense statistics can reveal aging patterns that are not well captured by summary statistics over atlas-defined regions as actual pattern boundaries might not well align with those regions. Weak findings reported by atlas-based statistics should therefore be pursued with the type of dense statistics proposed herein to take advantage of the complex information captured by structural MRI data sets.
(Support: AA021697; AA017168)

Imaging Methods:

Anatomical MRI

Modeling and Analysis Methods:

Image Registration and Computational Anatomy
Segmentation and Parcellation


Cortical Anatomy and Brain Mapping 1
Normal Development 2

Poster Session:

Poster Session - Tuesday


Statistical Methods

1|2Indicates the priority used for review

Would you accept an oral presentation if your abstract is selected for an oral session?


I would be willing to discuss my abstract with members of the press should my abstract be marked newsworthy:


Please indicate below if your study was a "resting state" or "task-activation” study.


By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute the presentation in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels or other electronic media and on the OHBM website.

I accept

Healthy subjects only or patients (note that patient studies may also involve healthy subjects):

Healthy subjects

Internal Review Board (IRB) or Animal Use and Care Committee (AUCC) Approval. Please indicate approval below. Please note: Failure to have IRB or AUCC approval, if applicable will lead to automatic rejection of abstract.

Yes, I have IRB or AUCC approval

Please indicate which methods were used in your research:

Structural MRI

For human MRI, what field strength scanner do you use?


Which processing packages did you use for your study?

Free Surfer
Other, Please list  -   Connectome Workbench

Provide references in author date format

[1] Dale AM, Fischl B, Sereno MI (1999), Cortical surface-based analysis. I. Segmentation and surface reconstruction, Neuroimage, 9(2):179-194.
[2] Desikan et al. (2006), An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest, NeuroImage, 31(3):968-80.
[3] Fischl et al. (2004), Automatically Parcellating the Human Cerebral Cortex, Cerebral Cortex, 14(1):11-22.
[4] Glasser MF, Van Essen DC. (2011), Mapping human cortical areas in vivo based on myelin content as revealed by T1- and T2-weighted MRI, Journal of Neuroscience, 31(32):11597-616.
[5] Glasser MF et al. (2013), The minimal preprocessing pipelines for the Human Connectome Project, Neuroimage, 80:105-24.
[6] Glasser MF et al. (2016), A multi-modal parcellation of human cerebral cortex, Nature, 536(7615):171-8.
[7] Pfefferbaum A et al. (2016), Adolescent Development of Cortical and White Matter Structure in the NCANDA Sample: Role of Sex, Ethnicity, Puberty, and Alcohol Drinking, Cerebral Cortex, 26(10):4101-21.
[8] Robinson EC et al. (2014), MSM: a new flexible framework for Multimodal Surface Matching, Neuroimage, 100:414-26.